CN103210646A - Method and apparatus for encoding/decoding images using the effective selection of an intra-prediction mode group - Google Patents

Abstract

One embodiment of the present invention relates to a method and apparatus for encoding/decoding images using the effective selection of an intra-prediction mode group. Said embodiment of the present invention relates to a method and apparatus for encoding/decoding images, which involve selecting an intra-prediction mode set using a pixel of a block adjacent to the current block, and generating a prediction block using one prediction mode of the selected intra-prediction mode set, thereby improving compression performance.

Description

Come method and apparatus to encoding/decoding image by effective selection intra prediction mode group

Technical field

The disclosure relates in some embodiments selects effective infra-frame prediction (intra-prediction) set of patterns to carry out the method and apparatus of encoding and decoding of video.More specifically, the disclosure relates to video coding/decoding method and device in some embodiments, described method and apparatus is selected the predictive mode collection based on the directivity of neighbor, makes the coding and the compression performance that omit for the additional information of selecting the predictive mode collection be improved.

Background technology

The statement of this part only provides and relates to background information of the present disclosure and may not constitute prior art.

Along with the development of the information and communication technology (ICT) that comprises the Internet, video communication increases outside text and voice communication.Traditional communication plan centered by text is difficult to satisfy consumer's various needs, thereby has increased the multimedia service that can comprise such as various types of information such as text, image, music.The multi-medium data amount is very big, thereby multi-medium data needs jumbo storage medium and needs to be used for the wide bandwidth that sends.Therefore, in order to send the multi-medium data that comprises text, image and voice data, may need a kind of compression coding scheme.

The basic principle of packed data is the processing that removes the data redundancy factor.When repeating, identical color or object can come packed data by removing spatial redundancy in piece image, when occurring variation seldom between the consecutive frame in frame of video or in audio signal, repeating identical note, can come packed data by removing time redundancy, consider that perhaps human vision and perception can come packed data by removing psycho-visual (psychovisual) redundancy to the insensitive situation of high-frequency.

In these video-frequency compression methods, industry has selected H.264 to carry out the AVC(advanced video coding), it has than MPEG-4(mpeg-4-) more excellent compression efficiency.As one of scheme that improves compression efficiency, H.264 user tropism's infra-frame prediction (hereinafter referred is infra-frame prediction) removes the chronotaxis in the frame.

Upside and leftward position and the current sub-block adjacent pixels of infra-frame prediction by being replicated in current sub-block in a predetermined direction predicted the value of current sub-block, and only current block pixel value and prediction piece difference between pixel values encoded.On the other hand, inter prediction (inter-prediction) (time prediction) is carried out prediction with reference to the scope of the frame that is positioned at last diverse location of time.Inter prediction is replenishing of inter prediction.For these two kinds of Forecasting Methodologies, more favourable a kind of method can replace another kind and be selected to image is encoded in essence.

According to the infra-frame prediction according to traditional H.264 standard, produce the prediction piece of current block based on another piece with coded sequence more early.Then, the value that obtains deduct the prediction piece from current block is encoded.For luminescence component, the prediction piece is produced by the piece of 4x4 or the macro block of 16x16.Piece for each 4x4 has nine kinds of selectable predictive modes, and for the piece of each 16,x16 four kinds of selectable predictive modes is arranged.In these predictive modes, be chosen in the predictive mode that causes minimum difference between current block and the prediction piece according to H.264 video encoder.

A plurality of selectable predictive mode collection rather than single collection may be arranged.What along with a plurality of predictive mode collection of selecting are provided, exist and the problem of concentrating selection relevant additional coding information at predictive mode.

Summary of the invention

Technical problem

An execution mode of the present disclosure that is used for addressing the above problem is by selecting the predictive mode collection and omitted the raising that compression performance is provided for the coding of the additional information of selecting the predictive mode collection based on the directivity of neighbor.

Technical scheme

Execution mode of the present disclosure provides a kind of encoding and decoding of video device, this encoding and decoding of video device comprises: video encoder, it is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, produce the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate, produce residual block by from described current block, deducting described prediction piece, produce residual block through transform and quantization by described residual block being carried out transform and quantization, and described residual block through transform and quantization is encoded; And Video Decoder, it is used for coming the residual block through transform and quantization is rebuild by the received code data, by the residual block through transform and quantization of rebuilding being carried out re-quantization and inversion brings the reconstruction residual block, the pixel of the adjacent block by use current block to be rebuild is selected the intra prediction mode collection, produce the prediction piece of described current block to be rebuild based on selected intra prediction mode collection, and rebuild described current block by the described prediction piece phase Calais of the residual block that will rebuild and described current block to be rebuild.

Another embodiment of the present disclosure provides a kind of video coding apparatus, this video coding apparatus comprises: intra predictor generator, it is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, and produces the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate; Subtracter, it is used for producing residual block by deduct described prediction piece from described current block; Converter and quantizer, it is used for producing residual block through transform and quantization by described residual block being carried out transform and quantization; And encoder, it is used for described residual block through transform and quantization is encoded.

Another execution mode of the present disclosure provides a kind of video decoder, and this video decoder comprises: decoder, and it is for the residual block of rebuilding by the received code data through transform and quantization; Inverse quantizer and inverse converter, it is used for by described residual block through transform and quantization being carried out re-quantization and inversion brings the reconstruction residual block; Intra predictor generator, it is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, and by using selected intra prediction mode collection to produce the prediction piece; And adder, it is used for rebuilding described current block in the Calais by the residual block of will rebuild mutually with described prediction piece.

Another execution mode of the present disclosure provides a kind of video coding/decoding method, this video coding/decoding method may further comprise the steps: the step of carrying out the video coding, this step is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, produce the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate, produce residual block by from described current block, deducting described prediction piece, produce residual block through transform and quantization by described residual block being carried out transform and quantization, and described residual block through transform and quantization is encoded; And the step of carrying out video decode, this step is used for coming the residual block through transform and quantization is rebuild by the received code data, by the residual block through transform and quantization of rebuilding being carried out re-quantization and inversion brings the reconstruction residual block, the pixel of the adjacent block by use current block to be rebuild is selected the intra prediction mode collection, produce the prediction piece of described current block to be rebuild based on selected intra prediction mode collection, and rebuild described current block by the described prediction piece phase Calais of the residual block that will rebuild and described current block to be rebuild.

Another embodiment of the present disclosure provides a kind of method for video coding, this method for video coding may further comprise the steps: the step of carrying out infra-frame prediction, this step is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, and produces the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate; Carry out the step of subtraction, this step is used for producing residual block by deduct described prediction piece from described current block; Carry out the step of transform and quantization, this step is used for producing residual block through transform and quantization by described residual block being carried out transform and quantization; And the step of carrying out coding, this step is used for described residual block through transform and quantization is encoded.

Another execution mode of the present disclosure provides a kind of video encoding/decoding method, and this video encoding/decoding method may further comprise the steps: carry out the step of decoding, this step is for the residual block of rebuilding by the received code data through transform and quantization; Carry out the step of re-quantization and inverse transformation, this step is used for by described residual block through transform and quantization being carried out re-quantization and inversion brings the reconstruction residual block; Carry out the step of infra-frame prediction, this step is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, and by using selected intra prediction mode collection to produce the prediction piece; And the step of carrying out addition, this step is used for rebuilding described current block in the Calais by the residual block of will rebuild mutually with described prediction piece.

Beneficial effect

According to the disclosure as described above, select the predictive mode collection based on the directivity of neighbor, be used for selecting the coding additional information of predictive mode collection to improve compression performance thereby therefore omitted.

Description of drawings

Fig. 1 is the block diagram that has schematically shown according to the video coding apparatus of one or more execution modes of the present disclosure;

Fig. 2 is the figure that shows the structure of intra predictor generator 110;

Fig. 3 be show size for the pixel of the current block of 4x4 (a～p) and with this current block adjacent pixels (figure of A～M);

Fig. 4 is the figure that shows intra prediction mode collection type;

Fig. 5 is the flow chart that shows the process of set of patterns selector 112 preference pattern collection;

Fig. 6 is the flow chart that shows according to the method for video coding of an execution mode of the present disclosure;

Fig. 7 is the block diagram that has schematically shown according to the structure of the video decoder of execution mode of the present disclosure;

Fig. 8 is the figure that shows the structure of intra predictor generator 730; And

Fig. 9 is the flow chart that shows according to the video encoding/decoding method of the disclosure one execution mode.

Embodiment

Hereinafter, will be described in detail with reference to the attached drawings execution mode of the present disclosure.In the following description, identical reference number indicates identical element, even they are shown in the different accompanying drawings.In addition, in below of the present disclosure, describing, for the sake of clarity, incorporate omission into the known function of this paper and the detailed description of structure.

In addition, when describing assembly of the present disclosure, can use such as first, second, A, B, (a) and (b) term.These terms only are the purposes of distinguishing mutually for an assembly and other assembly, rather than essence, order or the sequence of hint or suggestion assembly.If an assembly is described to " connection ", " coupling " or " link " to another assembly, then this may mean that these assemblies are not only directly " connection ", " coupling " or " link ", but also can be via third party's assembly " connection ", " coupling " or " link " indirectly.

Video coding apparatus and video decoder can be corresponding to such as the PC(personal computers), notebook, the PDA(personal digital assistant), the PMP(portable media player), the PSP(portable game machine), wireless communication terminal, smart phone, user terminals such as TV, can be corresponding to such as application server, server terminals such as service server, perhaps can be corresponding to various device, described various device comprises the communication equipment such as the communication modem that communicates with various device or wired, storage is used for image is encoded or decoded or is used for the various programs of interframe/infra-frame prediction and the memory of data are carried out in decoding or coding, be used for executive program and realize the microprocessor that calculates and control etc.

And, can pass through such as the Internet, Wireless Personal Network (WPAN), wireless lan (wlan), the WiBro(WiMAX, WiMAx just) network, wired such as mobile communications network or by such as cable, the USB(USB) various communication interfaces such as, send the image that is encoded into bit stream by video coding apparatus to video decoder in real time or in non real-time, and this bit stream is can be in video decoder decoded and can be resorted to video, and this video can be by playback then.

In general, video can be formed by a series of pictures, and each picture all can be divided into presumptive area, for example frame or piece.An image-region can be divided into piece, can be piece in the frame (intra-block) or interframe block (inter-block) with the block sort of dividing based on encoding scheme.Piece refers to the piece based on intraframe predictive coding scheme coding in the frame.The intraframe predictive coding scheme, and is encoded to the pixel difference of prediction piece and current block producing the prediction piece by the pixel using the pixel will carry out the piece that is encoded and decodes to be rebuild in the photo current of coding and predict current block.Interframe block is based on the inter prediction encoding scheme and the piece of encoding.The inter prediction encoding scheme is by predicting current block in the photo current producing the prediction piece with reference to the picture at least one past or the picture in a future, and the difference of this prediction piece and current block is encoded.Here, the frame of reference is called as reference frame when this current picture is encoded or decoded.

Fig. 1 is the block diagram that has schematically shown according to the video coding apparatus of one or more execution modes of the present disclosure.

Video coding apparatus 100 according to the disclosure one execution mode can comprise intra predictor generator 110, inter predictor 120, subtracter 130, converter and quantizer 140, encoder 150, inverse quantizer and inverse converter 160, adder 170 and frame memory 180.

It can be the unit input with the macro block that image encoded is carried out in expectation.In the disclosure, this macro block is the MxN form, and wherein each M and N all can have 2 ⁿSize, and M and N can be identical or differ from one another.Therefore, this macro block can be equal to or greater than macro block H.264.

The fallout predictor that is intra predictor generator 110 or inter predictor 120 can produce the prediction piece by the prediction current block.That is to say that fallout predictor 110 or 120 can be predicted the pixel value of each pixel of the current block that will encode in image, and can produce the prediction piece, described prediction piece has the predicted pixel values of each pixel.Here, this fallout predictor 110 or 120 can be predicted current block by the infra-frame prediction of intra predictor generator 110 execution or by the inter prediction that inter predictor 120 is carried out.

This inter predictor 120 can use different frame to produce the prediction piece with the prediction current macro.That is to say, inter predictor 120 can based on experienced that coding is handled and decoded previous frame in the pattern of inter predictor 120 produce motion vector by estimation, and in motion compensation process, use this motion vector to produce the prediction piece.

Intra predictor generator 110 is predicted the pixel of current block with the generation intra-frame prediction block by the pixel of using adjacent block, and produces the prediction piece by based on the correlation between the pixel of the correlation between the pixel of this adjacent block or intra-frame prediction block intra-frame prediction block optionally being carried out filtering.That is to say that intra predictor generator 110 can be by using encoded in the current macro and the neighbor of having rebuild, producing the prediction piece based on the pattern of intra predictor generator 110.

Subtracter 130 produces residual block by deduct the prediction piece from current block.That is to say, this subtracter 130 calculates the difference between pixel values of pixel value and the prediction piece that produces from intra predictor generator 110 or inter predictor 120 of each pixel of the current block that will encode, thereby produce this residual block, it has the residual signal of piece form.

Thereby the residual signal conversion that converter and quantizer 140 will produce from subtracter 130 also is quantified as a coefficient of frequency and produces residual block through transform and quantization.Here, suitable transform method can be the scheme that the picture signal in the spatial domain is converted to frequency-region signal, for example Hadamard transform or based on the integer transform (hereinafter referred is integer transform) of discrete cosine transform.Can use such as the DZUTQ(dead band and unify threshold value quantizing, hereinafter referred to as ' DZUDQ ') or quantize weighting matrix (Quantization Weighted Matrix) etc. as quantization scheme.

150 pairs of residual block by converter and quantizer 140 transform and quantizations of encoder encode to produce coded data.

Can adopt the entropy encoding scheme as this encoding scheme, but this is also not as restriction, and can uses various coding methods.

Required various information when in addition, encoder 150 can comprise the bit stream that obtains by the coefficient of frequency that quantizes is encoded and decode this coded bit stream in coded data.That is to say, coded data can comprise first field and second field, first field comprises the pattern by coding CBP(coded block, the block mode of coding) bit stream, Delta (delta) quantization parameter and the sampling frequency coefficient that obtains, second field comprises for the bit of the information of prediction (for example, at the intra prediction mode under the infra-frame prediction situation, motion vector etc. under the inter prediction situation) and other bits.

160 pairs of residual block through transform and quantization with quantizing by converter and quantizer 140 conversion of inverse quantizer and inverse converter are carried out re-quantization and inverse transformation, to rebuild residual block.Re-quantization and inverse transformation can be to be handled by the contrary of the performed transform and quantization of converter and quantizer 140.That is to say, inverse quantizer can be based on carrying out re-quantization and inverse transformation with the information relevant with transform and quantization (for example, the information of transform and quantization type) that quantizer 140 produces and transmits by oppositely carrying out by converter and quantizer 140 performed transform and quantization schemes from converter with inverse converter 160.

Adder 170 will be by rebuilding current block with the residual block phase Calais that is carried out re-quantization and inverse transformation by this inverse quantizer and inverse converter 160 by the prediction piece of fallout predictor 110 or 120 predictions.

The piece that frame memory 180 storages are rebuild by adder 170, and during infra-frame prediction or inter prediction, the piece of storing is produced the prediction piece as reference block.

Fig. 2 is the figure that shows the structure of intra predictor generator 110.

Intra predictor generator 110 is selected the intra prediction mode collection by the pixel of the adjacent block of use current block, and a predictive mode that adopts selected intra prediction mode to concentrate produces the prediction piece.

As shown in Figure 2, intra predictor generator 110 can be configured to comprise set of patterns selector 112 and prediction piece generator 114.

Fig. 3 be show size for the pixel of the current block of 4x4 (a～p) and with this current block adjacent pixels (figure of A～M).Fig. 4 is the figure that shows intra prediction mode collection type.

Set of patterns selector 112 uses the pixel of adjacent block to select intra prediction mode.Set of patterns selector 112 is selected the intra prediction mode collection based on the correlation between the pixel of this adjacent block.This correlation can be corresponding to the standard deviation between the pixel of adjacent block (standard deviation) or variance (variance), but the disclosure is not limited to this.

This set of patterns selector 112 can calculate correlation between the pixel of adjacent block based on the variance of equation 1 and equation 2.

${\mathrm{\σ}}_T=\underset{P=\{A,B,C,D\}}{\mathrm{\Σ}}{(P-{\mathrm{Mean}}_{A,B,C,D})}^2$ Equation 1

${\mathrm{\σ}}_L=\underset{P=\{I,J,K,L\}}{\mathrm{\Σ}}{(P-{\mathrm{Mean}}_{I,J,K,L})}^2$ Equation 2

(wherein P represents that each pixel and Mean represent mean value)

The variance that equation 1 calculates between the neighbor of current block upside, and the variance of equation 2 calculating between the neighbor in current block left side.

The variance equation of equation 1 or equation 2 can compare to determine whether to exist correlation by the threshold value (TH) with equation 3.

$\mathrm{TH}=\frac{{{\mathrm{<figure-callout\; id="2"\; label="Q\; step"\; filenames="HDA00003151094800022.png"\; state="\{\{state\}\}">Q</figure-callout>}}_{\mathrm{step}}}^2+8}{16}$ Equation 3

In the equation 3, Q _StepExpression quantization step parameter.

Fig. 5 is the flow chart that shows the processing of set of patterns selector 112 preference pattern collection.

As shown in Figure 5, judge (σ at step S502 _T＜TH) and (σ _L＜TH) whether these two conditions all satisfy, and when these two conditions all satisfy, select the fundamental mode collection at step S508.In the time can't satisfying these two conditions simultaneously, judge (σ at step S504 _T＜TH) whether this condition satisfies.As satisfied (σ _T＜TH) during this condition, select the horizontal pattern collection at step S512.When not satisfying (σ _T＜TH) during this condition, judge (σ at step S506 _L＜TH) whether this condition satisfies.As satisfied (σ _L＜TH) during this condition, select the vertical mode collection at step S510.When not satisfying (σ _L＜TH) during this condition, select the fundamental mode collection at step S508.

Prediction piece generator 114 uses the set of patterns of being selected by set of patterns selector 112 to produce the prediction piece.That is to say, by selecting to provide the predictive mode of optimum efficiency to produce the prediction piece in the predictive mode from be contained in selected set of patterns.

Fig. 6 is the flow chart that shows according to the method for video coding of one or more execution modes of the present disclosure.

As shown in Figure 6, method for video coding according to the disclosure one execution mode comprises set of patterns selection step S602, be used for selecting the intra prediction mode collection based on the correlation between the pixel of the adjacent block of current block, and prediction piece generation step S604, a predictive mode that is used for using selected intra prediction mode to concentrate produces the prediction piece.

Here, set of patterns selects step S602 corresponding to the operation of set of patterns selector 112, and predicts that piece produces step S604 corresponding to the operation of prediction piece generator 114, thereby detailed description will be omitted.

Fig. 7 is the block diagram that has schematically shown according to the structure of the video decoder of one or more execution modes of the present disclosure.

Video decoder 700 according to the disclosure one execution mode can be configured to comprise decoder 710, inverse quantizer and inverse converter 720, intra predictor generator 730, inter predictor 740, adder 750 and frame memory 760.

Decoder 710 is rebuild through the residual block of transform and quantization and the required information of decoding by the received code data.

Thereby decoder 710 can be decoded to coded data and be extracted the required information of piece decoding.The residual block that decoder 710 extracts in can first terminal from be contained in coded data and decoding has been encoded, can from second terminal that is contained in coded data, extract the information that forecasting institute needs, and the information that the forecasting institute that extracts can be needed passes to intra predictor generator 730 or inter predictor 740.

Thereby inverse quantizer and inverse converter 720 can carry out re-quantization and inverse transformation reconstruction residual block to decoded residual block through transform and quantization.

Fallout predictor (intra predictor generator 730 or inter predictor 740) produces the prediction piece by the prediction current block.In this example, Dui Ying fallout predictor 730 or 740 can adopt the mode identical with the fallout predictor (intra predictor generator 110 or inter predictor 120) of video coding apparatus 100 to predict current block.

Adder 750 will be by rebuilding current block in the Calais mutually with the prediction piece that is produced by fallout predictor 740 by the residual block that inverse quantizer and inverse converter 720 are rebuild.The current block of being rebuild by adder 750 can pass to frame memory 760, thereby can be used for predicting another piece in fallout predictor 730 or 740.

Frame memory 760 can be stored the image of reconstruction to be used for producing intra-frame prediction block and inter prediction piece.

Thereby decoder 710 can or extract through the residual block of transform and quantization and the required information of decoding the coded data decoding.The required information of decoding refers to be included in the required information of the coded bit stream of coded data and can be for decoding, for example, block type information, be the information of the intra prediction mode under the situation of intra prediction mode at predictive mode, be similar information in the information of information, transform and quantization type of the motion vector under the situation of inter-frame forecast mode and the various information at predictive mode.

Intra predictor generator 730 is selected the intra prediction mode collection by the pixel of the adjacent block of use current block, and uses selected intra prediction mode collection to produce the prediction piece.

Fig. 8 is the figure that shows the structure of intra predictor generator 730.

As shown in Figure 8, intra predictor generator 730 comprises set of patterns selector 732 and prediction piece generator 734.

Set of patterns selector 732 is selected the intra prediction mode collection based on the correlation between the pixel of the adjacent block of current block.The operation of the set of patterns selector 732 in the video decoder 700 can be identical or similar with the operation of set of patterns selector 112 in the video coding apparatus 100, thereby detailed description will be omitted.

Prediction piece generator 734 uses the set of patterns of being selected by set of patterns selector 732 to produce the prediction piece.That is to say, can be by selecting to provide the predictive mode of optimum efficiency to produce the prediction piece in the predictive mode from be contained in selected set of patterns.The operation of the prediction piece generator 734 in the video decoder 700 can be identical or similar with the operation of prediction piece generator 114 in the video coding apparatus 100, thereby detailed description will be omitted.

Fig. 9 is the flow chart that shows according to the video encoding/decoding method of one or more execution modes of the present disclosure.

As shown in Figure 9, the video encoding/decoding method according to the disclosure one execution mode comprises: set of patterns is selected step S902, and this step is used for selecting the intra prediction mode collection based on the correlation between the pixel of the adjacent block of current block; And prediction piece generation step S904, this step is used for producing the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate.

Here, set of patterns selects step S902 to produce step S904 corresponding to the operation of prediction piece generator 734 corresponding to the operation of set of patterns selector 732 and prediction piece, thereby detailed description will be omitted.

Can be by with the coded data output of the video coding apparatus 100 of Fig. 1 and the coding of the video decoder 700 of Fig. 7 according to the encoding and decoding of video device of the disclosure one execution mode; The data input connects to implement.

Encoding and decoding of video device according to the disclosure one execution mode comprises: video encoder, it is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, produce the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate, produce residual block by from described current block, deducting described prediction piece, produce residual block through transform and quantization by described residual block being carried out transform and quantization, and described residual block through transform and quantization is encoded; And Video Decoder, it is used for coming the residual block through transform and quantization is rebuild by the received code data, by the residual block through transform and quantization of rebuilding being carried out re-quantization and inversion brings the reconstruction residual block, the pixel of the adjacent block by use current block to be rebuild is selected the intra prediction mode collection, produce the prediction piece of described current block to be rebuild based on selected intra prediction mode collection, and rebuild described current block by the described prediction piece phase Calais of the residual block that will rebuild and described current block to be rebuild.

Here, video encoder can be by implementing according to the video coding apparatus 100 of the disclosure one execution mode, and Video Decoder can be implemented according to the video decoder 700 of the disclosure one execution mode.

Can implement by carrying out combination according to the method for video coding of the disclosure one execution mode and video encoding/decoding method according to the disclosure one execution mode according to the video coding/decoding method of the disclosure one execution mode.

Video coding/decoding method according to the disclosure one execution mode, this video coding/decoding method may further comprise the steps: the step of carrying out the video coding, this step is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, produce the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate, produce residual block by from described current block, deducting described prediction piece, produce residual block through transform and quantization by described residual block being carried out transform and quantization, and described residual block through transform and quantization is encoded; And the step of carrying out video decode, this step is used for coming the residual block through transform and quantization is rebuild by the received code data, by the residual block through transform and quantization of rebuilding being carried out re-quantization and inversion brings the reconstruction residual block, the pixel of the adjacent block by use current block to be rebuild is selected the intra prediction mode collection, produce the prediction piece of described current block to be rebuild based on selected intra prediction mode collection, and rebuild described current block by the described prediction piece phase Calais of the residual block that will rebuild and described current block to be rebuild.

In the above description, though all components of execution mode of the present disclosure all has been interpreted as being assembled into or being operably connected to the unit, the disclosure is not intended to and will self be limited to such execution mode.On the contrary, in target zone of the present disclosure, each assembly can be with quantity arbitrarily optionally and operationally combination.In these assemblies each can also realize with hardware independently that each assembly can optionally partly or as a whole combine simultaneously, and realizes to comprise for the computer program of the program module of carrying out the hardware equivalent functions.Those skilled in the art can easily derive code or the code segment that constitutes this program.Computer program can be stored in the computer-readable medium, and it can realize execution mode of the present disclosure when carrying out.Computer-readable medium comprises magnetic recording media, optical recording media and carrier media.

In addition, should be interpreted under default situations as the term of " comprising ", " comprising " and " having " and so on comprising property or open, rather than exclusiveness or sealing, unless carried out opposite definition clearly.All technology, science or other term is all consistent with the meaning understood by one of ordinary skill in the art, unless carried out opposite definition.Should in the context of correlation technique works, make an explanation to the generic term of finding in the dictionary, rather than too desirable or unrealistic, unless the disclosure has defined these terms clearly like this.

Though for exemplary purpose illustrative embodiments of the present disclosure is described, one of skill in the art will appreciate that under the situation that does not break away from inner characteristic of the present disclosure various modifications, increase and to substitute all be possible.Therefore, illustrative embodiments of the present disclosure is not to be described for the purpose that limits.Correspondingly, the scope of the present disclosure is not subjected to the restriction of above-mentioned execution mode, but is limited by claim and equivalent thereof.

Industrial applicibility

As previously shown, the disclosure is very useful in the application of video coding/decoding method and device, it is by selecting the predictive mode collection based on the directivity of neighbor, thereby is used for selecting the coding additional information of predictive mode collection to omit and compression performance can be improved.

The cross reference of related application

If can apply for, the application requires the priority at the patent application No.10-2010-0087387 of Korea S's submission on September 7th, 2010 based on 35U.S.C § 119 (a) so, and the whole contents of this korean application is incorporated this paper in the mode of quoting as proof.In addition, because identical reason, in the national requirements priority except the U.S., the whole contents of this korean application is incorporated this paper in the mode of quoting as proof to this non-provisional application based on this korean patent application.

Claims (18)

1. encoding and decoding of video device, this encoding and decoding of video device comprises:

Video encoder, it is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, produce the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate, produce residual block by from described current block, deducting described prediction piece, produce residual block through transform and quantization by described residual block being carried out transform and quantization, and described residual block through transform and quantization is encoded; And

Video Decoder, it is used for coming the residual block through transform and quantization is rebuild by the received code data, by the residual block through transform and quantization of rebuilding being carried out re-quantization and inversion brings the reconstruction residual block, the pixel of the adjacent block by use current block to be rebuild is selected the intra prediction mode collection, produce the prediction piece of described current block to be rebuild based on selected intra prediction mode collection, and rebuild described current block by the described prediction piece phase Calais of the residual block that will rebuild and described current block to be rebuild.

2. video coding apparatus, this video coding apparatus comprises:

Intra predictor generator, it is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, and produces the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate;

Subtracter, it is used for producing residual block by deduct described prediction piece from described current block;

Converter and quantizer, it is used for producing residual block through transform and quantization by described residual block being carried out transform and quantization; And

Encoder, it is used for described residual block through transform and quantization is encoded.

3. device according to claim 2, wherein said intra predictor generator comprises:

The set of patterns selector, it is used for selecting described intra prediction mode collection based on the correlation between the described pixel of the described adjacent block of described current block; And

Prediction piece generator, it is used for producing described prediction piece by a predictive mode that uses selected intra prediction mode to concentrate.

4. device according to claim 2, wherein said intra predictor generator is selected described intra prediction mode collection based on the correlation between the described pixel of described adjacent block.

5. device according to claim 4, wherein said correlation comprise standard deviation or the variance between the described pixel of described adjacent block.

6. video decoder, this video decoder comprises:

Decoder, it is for the residual block of rebuilding by the received code data through transform and quantization;

Inverse quantizer and inverse converter, it is used for by described residual block through transform and quantization being carried out re-quantization and inversion brings the reconstruction residual block;

Intra predictor generator, it is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, and by using selected intra prediction mode collection to produce the prediction piece; And

Adder, it is used for rebuilding described current block in the Calais by the residual block of will rebuild mutually with described prediction piece.

7. device according to claim 6, wherein said intra predictor generator comprises:

The set of patterns selector, it is used for selecting described intra prediction mode collection based on the correlation between the described pixel of the described adjacent block of described current block; And

Prediction piece generator, it is used for by using selected intra prediction mode collection and producing described prediction piece by the prediction mode information of described decoder reconstructs.

8. device according to claim 6, wherein said intra predictor generator is selected described intra prediction mode collection based on the correlation between the described pixel of described adjacent block.

9. device according to claim 8, wherein said correlation comprise standard deviation or the variance between the described pixel of described adjacent block.

10. video coding/decoding method, this video coding/decoding method may further comprise the steps:

Carry out the step of video coding, this step is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, produce the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate, produce residual block by from described current block, deducting described prediction piece, produce residual block through transform and quantization by described residual block being carried out transform and quantization, and described residual block through transform and quantization is encoded; And

Carry out the step of video decode, this step is used for coming the residual block through transform and quantization is rebuild by the received code data, by the residual block through transform and quantization of rebuilding being carried out re-quantization and inversion brings the reconstruction residual block, the pixel of the adjacent block by use current block to be rebuild is selected the intra prediction mode collection, produce the prediction piece of described current block to be rebuild based on selected intra prediction mode collection, and rebuild described current block by the described prediction piece phase Calais of the residual block that will rebuild and described current block to be rebuild.

11. a method for video coding, this method for video coding may further comprise the steps:

Carry out the step of infra-frame prediction, this step is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, and produces the prediction piece by a predictive mode that uses selected intra prediction mode to concentrate;

Carry out the step of subtraction, this step is used for producing residual block by deduct described prediction piece from described current block;

Carry out the step of transform and quantization, this step is used for producing residual block through transform and quantization by described residual block being carried out transform and quantization; And

Carry out the step of coding, this step is used for described residual block through transform and quantization is encoded.

12. method according to claim 11, the step of wherein carrying out described infra-frame prediction comprises:

The step that the execution pattern collection is selected, this step are used for selecting described intra prediction mode collection based on the correlation between the described pixel of the described adjacent block of described current block; And

Carry out the step that the prediction piece produces, this step is used for producing described prediction piece by a predictive mode that uses selected intra prediction mode to concentrate.

13. method according to claim 11, the step of wherein carrying out described infra-frame prediction comprises:

Select described intra prediction mode collection based on the correlation between the described pixel of described adjacent block.

14. method according to claim 13, wherein said correlation comprise standard deviation or variance between the described pixel of described adjacent block.

15. a video encoding/decoding method, this video encoding/decoding method may further comprise the steps:

Carry out the step of decoding, this step is for the residual block of rebuilding by the received code data through transform and quantization;

Carry out the step of re-quantization and inverse transformation, this step is used for by described residual block through transform and quantization being carried out re-quantization and inversion brings the reconstruction residual block;

Carry out the step of infra-frame prediction, this step is used for selecting the intra prediction mode collection by the pixel of the adjacent block that uses current block, and by using selected intra prediction mode collection to produce the prediction piece; And

Carry out the step of addition, this step is used for rebuilding described current block in the Calais by the residual block of will rebuild mutually with described prediction piece.

16. method according to claim 15, the step of wherein carrying out described infra-frame prediction comprises:

The step that the execution pattern collection is selected, this step are used for selecting described intra prediction mode collection based on the correlation between the described pixel of the described adjacent block of described current block; And

Carry out the step that the prediction piece produces, this step is used for by using selected intra prediction mode collection and producing described prediction piece in the prediction mode information that described decoding step is rebuild.

17. method according to claim 15, the step of wherein carrying out described infra-frame prediction comprises:

Select described intra prediction mode collection based on the correlation between the described pixel of described adjacent block.

18. method according to claim 17, wherein said correlation comprise standard deviation or variance between the described pixel of described adjacent block.

Images